COMPUTATIONALLY TRACTABLE ROBUST GNC FOR HYPERSONIC VEHICLES
Abstract
This proposed program research builds on an existing collaboration between PIs and BAE Systems through a sequence of Australian Research Council (ARC) Linkage grants from subsonic through to hypersonic vehicles. The joint effort successfully yielded a new methodology for a computationally effective optimization framework that combines CFD-based estimation of aerodynamics with the flight vehicle dynamic models and controller, to jointly maximize a ’global’ objective of performance requirements. The concurrent approach in deriving globally optimized solutions through co-design of both the control architecture and the aerodynamic structure, differentiates it from sequential design methods iterating hardware and software designs. This co-design philosophy was motivated by the opportunity it brings for risk reduction early in the design phase of complex platform solutions, but also enables a deeper understanding of the sources of performance limitations. The research objectives of this program will extend the work undertaken through the ARC collaborations with BAE Systems Australia to tackle the question of robustness in the presence of sources of uncertainty arising from moving targets; environmental disturbances and/or unmodeled dynamics leading to plant model-mismatch in the controller; thermal dynamics and management; and the underlying accuracy limitations of the CFD computations and resulting aero-decks. The emphasis of the proposed program subsequently shifts from near-optimal design to being able to guarantee mission completion in the presence of uncertainty, whilst also maintaining computational tractability in the process. From the aerodynamic structure perspective, this will mean consideration of thermal dynamics and the selective use of higher fidelity models (for example when active constraints are identified). For trajectory generation, this will mean augmentation of performance metrics with controllability measures to ensure that inner loops can respond appropriately to unmodeled disturbances. From the dynamic control perspective this will involve moving from a pre-determined fixed trajectory, to a series of backwards reachable sets that the vehicle must pass through to maximize mission outcomes in the presence of disturbances.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 20, 2023
- Source ID
- FA23862214074
Entities
People
- Chris Manzie
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Melbourne